Google Ads is the fastest way to put a service business in front of people searching for exactly what it sells. And the fastest way to spend a thousand dollars learning nothing. The difference is rarely bidding genius; it’s a handful of structural decisions made before launch and one weekly ritual after. This is the owner’s version: everything that matters, nothing that merely impresses.
Structure: small and legible
- Search campaigns first. Search captures existing demand. Someone typed the problem. Performance Max, Display, and video are for later, if ever; they spend fast and explain little at owner scale.
- One tight theme per ad group, few ad groups. A $30/day account with fifteen ad groups starves them all of data; three that each get real volume teach you something weekly.
- Phrase and exact match to start. Broad match hands the algorithm your wallet before you have conversion history to steer it; earn your way to broad, don’t start there.
- Geography honestly: target “presence” (people IN your area), not the default that includes people merely interested in it. The single most common silent budget leak for local businesses.
Budget: arithmetic before hope
Work backward from click prices, not forward from what feels affordable: budget ÷ CPC = clicks; clicks × conversion rate = leads. If your industry’s CPC makes your budget good for one click a day, the money is better spent elsewhere until it isn’t. Run the budget planner and the break-even calculator before committing, and give any test 30 days of consistent spend before judging it.
The ads themselves
Write for the search, not the brand: the headline echoes what they typed (this chain continues onto the page. message match), one concrete differentiator, one plain call to action. Responsive search ads take up to fifteen headlines; give the algorithm genuinely different angles to test, not fifteen rephrasings. Character limits and live counts: the ad-copy checker.
The weekly ritual that does the optimizing
Skip the dashboard tourism. One ritual, twenty minutes, weekly: read the search-terms report. The actual searches that triggered your ads. Add negatives for the junk, and note which real searches convert. Over a few months this compounds into a moat: a negative list and keyword set tuned to your exact market that no competitor can copy.
What to ignore
Auto-applied recommendations (review them; most serve Google’s revenue, not yours. Especially “switch to broad match”), optimization score (a sales metric wearing a quality costume), and any agency selling “we’ll manage it for 10% of spend” without mentioning conversion tracking in the first conversation. The work above is boring and sufficient.